Families of victims of the February school shooting in Tumbler Ridge, B.C., sued OpenAI in U.S. federal court, alleging negligence for failing to alert police. OpenAI says it flagged and banned the shooter's ChatGPT account last June for discussing violence but judged it below the threshold for an RCMP referral; Sam Altman has since apologised.
New Orleans police appear to still be using Project NOLA's private live face recognition network despite a city law and a claimed pause, according to emails obtained by a privacy researcher showing officers sending photos to it as of November 2025. NOPD dismissed the researcher's complaints, claiming the requests weren't a 'use' of the technology.
Magistrate Judge Peter H. Kang of the Northern District of California fined Fresno attorney Lenden Webb $1,001 and ordered four hours of continuing legal education for failing to supervise Katherine Cervantes, a lawyer at Webb's firm, after the pair submitted a brief in Hill v. Workday citing nonexistent law, partly due to AI. Webb says the issue 'is behind us.'
Magistrate Judge Gabriel W. Gorenstein of the Southern District of New York fined attorney Tricia S. Lindsay $2,500 under Rule 11 for briefs citing nonexistent cases, finding an AI program generated them and not crediting Lindsay's hint that LexisNexis was responsible. Lindsay must also send the order to the client and alert judges in all pending cases.
Police in Durg, Chhattisgarh, registered an FIR against unidentified persons over a suspected AI-generated deepfake video targeting former chief minister Bhupesh Baghel, allegedly shared from two Instagram handles, and asked Instagram for the handler's details. Congress leaders call the clip fake and defamatory; the state women's commission has sought strict action.
Scammers are using AI-generated videos of celebrities including Taylor Swift and Rihanna to promote rewards schemes on TikTok, according to authentication company Copyleaks. The ads, some bearing TikTok branding, redirect users to third-party services asking for personal information; in one, an AI Swift urges sign-ups to a feature called TikTok Pay.
The House Homeland Security and China Select committees sent letters on Wednesday to the CEOs of Airbnb and Cursor maker Anysphere seeking information on their use of Chinese AI models, according to documents viewed by Semafor. Anysphere disclosed its Composer 2 model was built on Moonshot AI's Kimi; Airbnb used Alibaba's Qwen for a customer service agent.
ReversingLabs says a malicious npm package, @validate-sdk/v2, disguised as a validation tool, was added to an autonomous trading agent in February 2026 in a commit reportedly co-authored by Anthropic's Claude Opus. Attribution for the PromptMink campaign, spanning more than 60 packages, points to North Korean state-sponsored group Famous Chollima; the malware steals secrets and exposes crypto wallets.
